The company deploys an artificial intelligence (AI) agent. The agent makes a mistake. The insurance policy does not cover this.
This result is no longer hypothetical. Major insurance companies are excluding AI from standard coverage for businesses. State regulators are approving the applications and a new market is already forming to fill the gap.
Carriers point out that they cannot price the risk
Berkshire Hathaway, Chubb and Travelers It sought approval from government regulators to exclude AI-related damages from general liability policies. State regulators approved more than 80% of those requests I mentioned. Florida, Connecticut and Maryland approved the largest number of applications. The exceptions went into effect early in January.
The moves follow two new voluntary approvals introduced by the Insurance Services Bureau, a private body that sets industry standards. Policyholder pulse I mentioned Some carriers, including Berkeley, have introduced absolute exclusions for AI across their directors and officers, errors and omissions, and fiduciary liability policies. The endorsement covers bodily injury, property damage, and personal and advertising injuries associated with generative AI output, including defamatory content, intellectual property infringement, and physical damages traceable to errors caused by the AI.
BHL Company male ISO forms support approximately 82% of property and casualty policies in the United States. Adoption is expected to be rapid. Many carriers will attach these endorsements when you renew.
What falls outside the scope of coverage now
Plaintiffs have presented a wide range of legal theories in filings related to artificial intelligence. Policyholder pulse Indexed Categories: Copyright and intellectual property claims arising from training large language models, privacy and data use claims, antitrust claims, discrimination and algorithmic bias claims, and securities class actions related to artificial intelligence. Each now falls into a grayer area for companies whose carriers have secured AI exemptions.
Advertisement: Scroll to continue
Insurers are also moving to limit AI losses in cybersecurity policies, the Financial Times reported. This narrows down the options across multiple policy types at once. Policyholder pulse It has been marked The courts have not yet settled on the scope of application of these exceptions and that the effect for some policyholders may make their coverage illusory. BHL Company Found Small and medium-sized companies, which often do not have specialized coverage, face the greatest exposure.
A new market fills the gap
Specialized companies intervene. Munich RE And startups including Corgi, jacket, Mayflower Specialty and Embroidered We now offer independent AI liability policies, Information I mentioned. Coverage limits range from $2 million to $50 million. Premiums range from a few hundred dollars to several hundred thousand dollars annually.
This pattern mirrors how insurance companies approached cybersecurity a decade ago. A wave of attacks has sparked corporate claims. Companies have successfully claimed that traditional policies cover losses because the policies did not explicitly exclude them. Insurers have taken the Internet out of standard coverage and built standalone products. That market has matured. Now, the AI liability market has begun the same process.
The tension lies within the industry itself. piments I mentioned Insurance giants are deploying AI agents to coordinate the entire workflow across claims, underwriting and policy servicing. The carriers themselves are simultaneously withdrawing AI coverage from the policies they sell. PYMNTS too I mentioned AI is beginning to transform underwriting itself, compressing timelines and changing how risk is priced. Insurers are betting on AI internally while refusing to internalize AI risks externally.
BHL Company It is recommended Companies should review existing policies for new endorsements with their brokers and consider seeking affirmative AI coverage through technology errors and omissions policies, cyber liability insurance, or emerging standalone AI products. Strengthening internal AI governance, conducting bias testing, and disclosing the use of AI can also reduce exposure.
Insurers are not saying that AI is uninsurable. They say they don’t know yet what the cost will be. Until they do, the risk falls on the companies deploying this technology.
